mahsa's profile - activity

I want to sum all values in 4 frames in real-time. current frame and 3 frames before that. To do that, every time that current farme is produced, I store it in a vector called Nframe. This is what I have done until now but seems incorrect..

float *amp;   //points to the current frame
vector<float *>Nframe;

   for(int i=0; i<4; i++){
           Nframe.push_back(amp);
        }

         int sum_of_elems=0;
         for(std::vector<int>::iterator j=Nframe.begin();j!=Nframe.end();++j){
         sum_of_elems += *j;
         }

I could detect a shape by using canny edge detector. is there any way to find the center point of this detected object? (I mean is there any function in Opencv that can get the output of canny edge detector function as an input and then give the center of this area as an output?)

Thanks in advance..

Thanks for the answer...One more thing: Type of the input is CV_32FC1, should I still use uchar?

I was searching for standard deviation calculation using opencv that I found this:

http://answers.opencv.org/question/6501/problem-with-meanstddev/

I can not underestand what do these two lines mean:

uchar       mean_pxl = mean.val[0];
uchar       stddev_pxl = stddev.val[0];

what would be the output of mean.val[0] and stddev.val[0] ?

Thanks in advance..

I have a matrix which has values between 100 to 900. I want to just keep values between 300 to 400 and make zero rest of them. I could not find any thresholding function which keep values between a Range. Does Opencv have any function to do that?

Thanks in advance...

wow..good news. Is there any function for "orientation"?

I have used sobel to get the gradient of an image in X and Y direction:

cv::Sobel( Min1, grad_x, Min1.depth(), 1, 0, 3);    
cv::Sobel( Min1, grad_y, Min1.depth(), 0, 1, 3);

now I have the gradient of the Min1 in X direction stored in Mat grade_x and in Y direction stored in Mat grade_y.

How do I calculate the magnitude of grad_x and grad_y? Is there any function in Opencv to get the norm of a matrix?

Thanks in advance...

you meant I should try to find the gradient in X and Y direction and use tan to get the orientation?

wow..Gradient Images, sounds interesting..do you have any good book or website which is relevant to it?

In spatial domain, Yes, it is possible but I need to have orientation and magnitude as output.. which is impossibel in spatial domain

Let me to explain what I meant more..assume that I am interested in output(magn and orientation) of the pixel which is located in row 4 and column 8. Is it meaningful to just calculate DFT for this specific pixel? In DFT calculation seems it is possible but in Image procesing viewpoint...I do not know! do you think neighbour pixels have any affection on the output of each pixel value at the end?

Thanks for the clear answer...I came up with a new question: If I define an empty kernel on the image, is there any difference If I apply DFT on whole image or just apply it into this specific kernel?I mean what would be the output of pixel values inside the kernel when I apply DFT into whole image and when I apply it on this kernel?In better way: I want to get magnitude and orientation of some values inside the imege. Should I apply DFT on whole image and then find the correspond values result or should I try to calculate DFT on those specific values(for example by assuming that they are defined inside a kernel)? It would be appreciated if you make it clear for me...

I used DFT in opencv to get the complex and real images values but some of the values which I get for the Complex image are Negative. Is it usual? what is the reason that I get Negative values?

Thanks in advance..

because each element of the list is a float pointer to the current frame.

Is there any possibility in opencv to get average of some frames?

std::list<float*> frames;

for(int i = 0; i < frames.size(); i++)
 {
     //Average of frames            

 }

Thanks a lot for your help...what I am doing is difficult for me since I am completely newbie. I edited the code, it would be appreciated if you give me some hints on what I am doing..

thanks for the comment but what I mean is using some kind of Temporal Filtering..I want to use medianBlur function in temporal domain

I want to apply medianblur function into a sequence of frames...So, is it possible two add some frames the same as yours, then get the average of them and give it as an input to the medianBlur function?

really? so, what is addweighted function?!!

is there any way to add several images two each other using opencv?

thanks in advance...

after editing:

 int N;
     cin >> N;
     for(int i=0; i<sen; i++){
        Nframes.push_back(pClose); //pclose is a float pointer which points to the current frame
        if(Nframes.size() == N){
            for(int i=N; i>0; i--){
               Nframes.pop_front();

             //  addWeighted(src1, alpha, src2, beta, gamma, dst, -1);

           }
        }
     }

actually what I am going to do is noise reduction. I am using depth images and I want to work on motion blur. I do not know what would be the benefit of tracking 1 point in stream and what would be the result of getting average of a sequence of frames..somehow applying a kind of filtering algorithm into a sequence of frames. would it smooth the images?

Thanks Median, but I am talking about the real-time...Is that work in this case since I just have access to the current frame?

It is the real one..thanks for the comment.. actually the problem was thresholding part...

I would like to ask a question which I could not find any answer for that in the internet. I was thinking about the Temporal filtering that I came up with this question that If it is possible to get average of some sequence of frames or not? if yes, what would be the benefit of doing this? Does it smooth the images? Is it a kind of temporal filtering? and....

Any explanation would be appreciated...

Is it possible to add N frames together in Real-time by using Opencv?

I have one question! What is the reason that My HoughCircles function sometimes can detect the ball sometimes not?

there is just one ball in the scene, without any movement... sometimes HoughCircles can detect it sometimes not!

any idea will be appreciated..

If I want to draw a line between (center.x , center.y)in the current frame and (center.x , center.y) in the previous frame..should I every time that the code runs keep it in a vector and then take them two by two and use cvPoint() and cvLine() ?

Is this the correct way to do that by using opencv?

If I know the coordinate of one point in Frame 1, Is there any way in Opencv to track this point in a sequence of frames and find its new coordinates in different frames?Is there any way to Plot them?

Thanks in advance